The central purpose of this study is to estimate the effects of alcohol taxes on consumption and traffic fatalities. Our approach improves on most previous work in three ways. First, we carefully examine the available data on taxes and beverage prices. We assess their reliability by comparing them to detailed Consumer Price Indices for beer, wine, and spirits, and by performing formal statistical tests. Second, we estimate separate relationships between taxes and prices, prices and consumption, and consumption and fatalities. Specifically, we estimate the extent to which alcohol taxes are passed forward to beverage prices, the extent to which consumers respond by decreasing consumption, and the effects of reduced consumption on fatalities. Third, we estimate the degree to which increases in the price of one beverage cause consumers to substitute other beverages, and thus, the effectiveness of tax increases in reducing total alcohol consumption. For example, increased spirits taxes may cause consumers to increase their consumption of beer or wine.